Workflow: Extract Classified Point Cloud Regions

Use the Extract Classified Point Cloud Regions command to automatically extract from a selected point cloud region ASPRS (American Society for Photogrammetry and Remote Sensing) classes into individual point cloud regions.

You can select to show or hide (using the View Filter Manager), or export, any point cloud region in your project, including classified point cloud regions.

Following are the basic steps for this workflow. For more detailed instructions, see Extract Classified Point Cloud Regions.

Notes...

All workflows...


Steps:		Commands:
1.	Display the Extract Classified Point Cloud Regions command pane.	Extract Classified Point Cloud
2.	In the Classification type drop-down list, select the type of classification you want to perform. Classification types... Outdoor (Aerial): This classification type is based on deep learning technology to provide overall more accurate aerial extractions and classifications. Ground - Extracts a single Ground region from the entire ground identified in the selected scan points. Ground regions are colored brown by default. The ASPRS classification is 2. Buildings - Extracts a separate Building region from each building identified in the selected scan points, or, optionally, a single Building region from all buildings. If more than one Building region is extracted, each subsequent region name is appended with a sequential number (for example, Building, Building_1, Building_2, and so on). Building regions are colored blue by default. The ASPRS classification is 6. Poles and Signs - Extracts a separate Poles and Signs region from each pole and sign identified in the selected scan points, or, optionally, a single Poles and signs region from all poles and signs. If more than one Poles and Signs region is extracted, each subsequent region name is appended with a sequential number (for example, Poles and signs, Poles and Signs_1, Poles and Signs_2, and so on). Poles and Signs regions are colored red by default. The ASPRS classification is 65. Medium Vegetation - Extracts a single Medium Vegetation region from all of the medium vegetation (for example, shrubs) identified in the selected scan points. Medium Vegetation regions are colored light green by default. The ASPRS classification is 4. High Vegetation - Extracts a single High Vegetation region from all of the tall vegetation (for example, trees) identified in the selected scan points. High Vegetation regions are colored green by default. The ASPRS classification is 5. Power Lines - Extracts a single Power Lines region from all of the power lines identified in the selected scan points. Power Lines regions are colored yellow by default. The ASPRS classification is 14. Low point (noise) - The ASPRS classification is 7. Dividers - Extracts a single Dividers region from all dividers (for example, fences) identified in the selected scan points. Dividers regions are colored light blue by default. Indoor: Floor - Extracts a man-made flat surface accessible by a walking person. In industrial environments, it is usually concrete. In the case of multiple-floors, only the lowest one will be extracted. Wall - Extracts vertical structures that enclose a space as part of the building envelope. The extraction is intended to work only with vertical walls (flat or curved). Inclined walls are not extracted. Grated Floor - Extracts a man-made floor composed of a regularly spaced collection of essentially identical, parallel, elongated elements. In an industrial environment, grated floors are typically metallic and prevent access through an opening while permitting communication or ventilation. Ceiling - Extracts an upper interior surface of a room or other similar compartment. The extraction is intended to work only with a flat ceiling. Both grating and concrete ceilings will be extracted. Outdoor (Terrestrial): This classification type is based on deep learning technology to provide overall more accurate terrestrial extractions and classifications. Ground - Extracts a single Ground region from the entire ground identified in the selected scan points. Ground regions are colored brown by default. The ASPRS classification is 2. Steps - Extracts a single Steps region from all of the steps identified in the selected scan points. Steps regions are colored gray by default. The ASPRS custom classification is 73. Medium Vegetation - Extracts a single Medium Vegetation region from all of the medium vegetation (for example, shrubs) identified in the selected scan points. Medium Vegetation regions are colored light green by default. The ASPRS classification is 4. High Vegetation - Extracts a single High Vegetation region from all of the tall vegetation (for example, trees) identified in the selected scan points. High Vegetation regions are colored dark green by default. The ASPRS classification is 5. Low point (noise) - The ASPRS classification is 7. Buildings - Extracts a single Building region from all buildings identified in the selected scan points. Building regions are colored blue by default. The ASPRS classification is 6. Dividers - Extracts a single Dividers region from all dividers (for example, fences) identified in the selected scan points. Dividers regions are colored light blue by default. The ASPRS custom classification is 79. Poles - Extracts a single Poles region from all poles identified in the selected scan points. Poles regions are colored red by default. The ASPRS classification is 65. Signs - Extracts a single Signs region from all signs identified in the selected scan points. Signs regions are colored cyan by default. The ASPRS classification is 66. Power Lines - Extracts a single Power Lines region from all of the power lines identified in the selected scan points. Power Lines regions are colored yellow by default. The ASPRS classification is 14. Railway: This classification type is based on deep learning technology to provide overall more accurate railway extractions and classifications. Rail - Extracts a single Rail region from all railway rails identified in the selected scan points. Rail regions are colored red by default. The ASPRS classification is 16. Sleeper - Extracts a single Sleeper region from all railway sleepers identified in the selected scan points. Sleeper regions are colored blue by default. The ASPRS custom classification is 240. Ballast - Extracts a single Ballast region from all railway ballast identified in the selected scan points. Ballast regions are colored tan by default. The ASPRS custom classification is 241.
3.	Select the point cloud region(s) from which you want to extract classified regions.
4.	In the Classes to extract list, select each class you want to extract.
5.	If you selected the Indoor classification type, optionally specify that only a single region be created for a classification when multiple features are found (instead of one region per feature) by checking the Create one new region per class check box.
6.	If you selected the Railway classification type with Ballast extraction, do either of the following in the Advanced Settings drop-down section: To use offset distances from a trajectory to determine the area to be searched for Ballast extraction/classification, click in the Limit classification area by trajectory field, and select an existing trajectory or linestring. Then, select the From trajectory option and enter the Balast distance from the selected trajectory/linestring to specify the area to be searched. To use offset distances from the rails to determine the area to be searched for Ballast extraction/classification, select the From rail option. Then, enter the Ballast distance from each of the two rails to specify the area to be searched.
7.	Click the Classify button.
8.	If you selected the Outdoor (Terrestrial) or Outdoor (Aerial) classification type and you are not satisfied with the classifications that were performed, in the Refine Classification list, check the check box for each extracted class type you want to refine. Then click the Refine button.
9.	Optionally, use the View Filter Manager to hide or show any point cloud regions.	View Filter Manager

All workflows...